In earlier work, we developed an R-C model which enabled us to predict calorimeter performance within 5%. The model revealed several sources of error in the current calorimeter design. With this information, we re-designed the calorimeter to reduce these errors and to offer increased performance and flexibility. The re-design (a) eliminates the air gap, (b) requires only one sensor, (c) can accept all three currently used cell sizes 0.3, 0.5 and 1.0 ml, and (d) increases sensitivity with no significant change in rise-time. During FY83, this design was implemented and evaluated. All of the above features were realized. However, with the larger cells, the magnitude of the mixing artifact increased to the point that it negated the advantage of using the large cell volume. With the large 1 ml cell, the mixing artifact due to the rotation of the cell averaged 900 micro-joules with a large variability of (plus and minus)700 micro-joules. The major cause of this artifact was found to be a small temperature gradient (4m C) within the constant temperature chamber. The larger cell volumes were much more sensitive to this gradient than the smaller cells. The gradient was reduced to less than 0.5m C by shunting different currents through the heating pads. This reduced the artifact to approximately 100(plus and minus)30 micro-joules.